In 2002, our group was asked by NAS/NRC to review all evidence linking water quality indicators to human health. We published these results as a meta-analysis (Wade et al. 2003). Traditional indicators (total coliform, fecal coliform, and Enterococcus) are problematic because prolonged analysis times (18-96 hours) make timely warnings difficult. Advances in microbiologic techniques are now pronounced and justify a comparison of current indicators with more rapidly measured new indicators. Additionally, important advances in statistical methods (such as the estimation of multivariate dose-response functions using inverse-probability of exposure weighting) have not been used to investigate the possibility of important multivariate relationships of these indicators to health and prediction. As preparation, we conducted a study at 6 San Diego beaches in 2003. We enrolled 12,458 beach users, recorded detailed swimming histories and collected illness data for 14 days. We collected hourly water samples at sites linked to the swimmer locations. We present analyses of this pilot data showing our ability to detect risk from swimming and a strong association with health outcomes of 1 novel indicator (F+ coliphage) used in the pilot. We present newly released EPA data showing a strong correlation between gastrointestinal (Gl) illness and Enterococcus measured by rapid QPCR (same method we use) in a freshwater site. We propose, over 4 summers, to enroll at least 17,600 beach users at Doheny Beach (one of California's most contaminated public beaches) using the same design and expanding our indicator testing to include rapid methods and viral agents from water samples spatially linked to each swimmer. Our 2003 pilot study research team is entirely intact (UC Berkeley, SCCWRP, Harvard, UNO, OHSU) and now has been joined by the NOAA Hollings Marine Lab. Decisions about beach closures have enormous economic implications. If successful, we will quantify the relationship between health and novel rapid methods for traditional indicators as well as viral indicators.